Search

Abstract Background The determinants of household food insecurity (HFI) do not act in isolation, and are known to be complex, stochastic, nonlinear, and multidimensional. Despite this being especially true in periods of shocks, studies that focus on integrated modelling of the HFI determinants during the COVID-19 lockdown are scarce, with no available evidence on Uganda. The main objective of this study was to develop Bayesian belief network (BBN) models to analyse, rank, and illustrate the conceptual reasoning, and complex causal relationships among the determinants of HFI during the COVID-19 lockdown. This study was based on seven rounds of Uganda’s High-Frequency Phone Surveys data sets collected during the lockdown. A total of 15,032 households, 17 independent determinants of HFI, and 8 food security indicators were used in this study. Metrics of sensitivity, and prediction performance were used to evaluate models’ accuracy. Results Eight BBN models were developed for each food insecurity indicator. The accuracy rates of the models ranged between 70.5% and 93.5%, with an average accuracy rate of 78.5%, indicating excellent predictive performance in identifying the determinants of HFI correctly. Our results revealed that approximately 42.2% of the sampled households (n = 15,032) in Uganda were worried about not having enough food. An estimated 25.2% of the respondents reported skipping a meal, while 32.1% reported consuming less food. Less than 20% of the households experienced food shortage, hunger, or having nothing to eat. Overall, 30.6% of the households were food insecure during the lockdown. The top five ranked determinants of HFI were identified as follows: (1) households’ inability to produce enough food; (2) households’ inability to buy food; (3) reduced household income; (4) limited cash assistance, and (5) households’ inability to stock adequate food supplies. Conclusions Ranking, rather than the statistical significance of the determinants of HFI, is crucial as an approach to applied research, as it helps stakeholders determine how to allocate resources for targeted interventions within the constraints of limited funding. These findings emphasize the importance of intervening on the most highly ranked determinants of HFI to enhance the resilience of local food systems, and households’ capacity to cope with recurring and unforeseen shocks.

Abstract Background Malaria risk factors at household level are known to be complex, uncertain, stochastic, nonlinear, and multidimensional. The interplay among these factors, makes targeted interventions, and resource allocation for malaria control challenging. However, few studies have demonstrated malaria’s transmission complexity, control, and integrated modelling, with no available evidence on Uganda’s refugee settlements. Using the 2018–2019 Uganda’s Malaria Indicator Survey (UMIS) data, an alternative Bayesian belief network (BBN) modelling approach was used to analyse, predict, rank and illustrate the conceptual reasoning, and complex causal relationships among the risk factors for malaria infections among children under-five in refugee settlements of Uganda. Methods In the UMIS, household level information was obtained using standardized questionnaires, and a total of 675 children under 5 years were tested for malaria. From the dataset, a casefile containing malaria test results, demographic, social-economic and environmental information was created. The casefile was divided into a training (80%, n = 540) and testing (20%, n = 135) datasets. The training dataset was used to develop the BBN model following well established guidelines. The testing dataset was used to evaluate model performance. Results Model accuracy was 91.11% with an area under the receiver-operating characteristic curve of 0.95. The model’s spherical payoff was 0.91, with the logarithmic, and quadratic losses of 0.36, and 0.16 respectively, indicating a strong predictive, and classification ability of the model. The probability of refugee children testing positive, and negative for malaria was 48.1% and 51.9% respectively. The top ranked malaria risk factors based on the sensitivity analysis included: (1) age of child; (2) roof materials (i.e., thatch roofs); (3) wall materials (i.e., poles with mud and thatch walls); (4) whether children sleep under insecticide-treated nets; 5) type of toilet facility used (i.e., no toilet facility, and pit latrines with slabs); (6) walk time distance to water sources (between 0 and 10 min); (7) drinking water sources (i.e., open water sources, and piped water on premises). Conclusion Ranking, rather than the statistical significance of the malaria risk factors, is crucial as an approach to applied research, as it helps stakeholders determine how to allocate resources for targeted malaria interventions within the constraints of limited funding in the refugee settlements.

Abstract Background While 5% of 247 million global malaria cases are reported in Uganda, it is also a top refugee hosting country in Africa, with over 1.36 million refugees. Despite malaria being an emerging challenge for humanitarian response in refugee settlements, little is known about its risk factors. This study aimed to investigate the risk factors for malaria infections among children under 5 years of age in refugee settlements in Uganda. Methods We utilized data from Uganda’s Malaria Indicator Survey which was conducted between December 2018 and February 2019 at the peak of malaria season. In this national survey, household level information was obtained using standardized questionnaires and a total of 7787 children under 5 years of age were tested for malaria using mainly the rapid diagnostic test. We focused on 675 malaria tested children under five in refugee settlements located in Yumbe, Arua, Adjumani, Moyo, Lamwo, Kiryadongo, Kyegegwa, Kamwenge and Isingiro districts. The extracted variables included prevalence of malaria, demographic, social-economic and environmental information. Multivariable logistic regression was used to identify and define the malaria associated risk factors. Results Overall, malaria prevalence in all refugee settlements across the nine hosting districts was 36.6%. Malaria infections were higher in refugee settlements located in Isingiro (98.7%), Kyegegwa (58.6%) and Arua (57.4%) districts. Several risk factors were significantly associated with acquisition of malaria including fetching water from open water sources [adjusted odds ratio (aOR) = 1.22, 95% CI: 0.08–0.59, P = 0.002], boreholes (aOR = 2.11, 95% CI: 0.91–4.89, P = 0.018) and water tanks (aOR = 4.47, 95% CI: 1.67–11.9, P = 0.002). Other factors included pit-latrines (aOR = 1.48, 95% CI: 1.03–2.13, P = 0.033), open defecation (aOR = 3.29, 95% CI: 1.54–7.05, P = 0.002), lack of insecticide treated bed nets (aOR = 1.15, 95% CI: 0.43–3.13, P = 0.003) and knowledge on the causes of malaria (aOR = 1.09, 95% CI: 0.79–1.51, P = 0.005). Conclusions The persistence of the malaria infections were mainly due to open water sources, poor hygiene, and lack of preventive measures that enhanced mosquito survival and infection. Malaria elimination in refugee settlements requires an integrated control approach that combines environmental management with other complementary measures like insecticide treated bed nets, indoor residual spraying and awareness. Graphical Abstract

Smallholder farmers in sub-Saharan Africa are at a greater risk to the impacts of climate variability. We therefore sought to assess vulnerability of smallholder sorghum farmers to climate variability in Kigezi highlands of south-western Uganda. A vulnerability index that integrates selected socio-economic and biophysical variables was obtained through key informant interviews and household surveys, from 230 conveniently sampled sorghum farming households within three sub-counties differentiated by altitude. Rainfall data were obtained from Uganda National Meteorological Authority. Quantitative data were analysed using Statistical Package for Social Sciences (version 23) and STATA software to generate inferential and descriptive statistics, notably frequencies, percentages and chi-square tests, to establish relationship between variables. Content analysis was used to generate themes emerging from the qualitative data. The overall vulnerability index results indicate Kashambya as the most vulnerable (6.9), followed by Bubare (1.8), while Kamwezi was the least vulnerable (–0.2). This study recommends targeted extension services such as access to customised weather information and better agronomic practices to reduce smallholder sorghum farmers’ vulnerability.